RF components, such as amplifiers, filters, switches, splitters, and couplers, have a frequency dependent transfer function. Even when the desire is to achieve a flat transfer function (no gain or phase variations over frequency), RF components often introduce variations that detract from their performance of their circuits. As a result, the RF circuits of RF components often exhibit gain or phase variations over frequency, which results in a non-flat RF output signal, despite of the desire to emit a “flat” response across frequency. In some cases, it is desirable to emit a certain profile of RF power vs. frequency, and the RF circuit and/or RF components imperfections cause some variations relative to the desired profile.
The transfer function of RF components (especially active devices such as amplifiers) is also affected by temperature. Both the nominal gain and the gain function over frequency are typically affected by temperature. Thus, the output of an RF device typically departs from the desired gain vs. frequency profile to a greater degree than normal when the temperature varies, and especially when the temperature reaches the FR device's intended temperature operation extremes.
Often, the imperfections introduced by the RF component and its RF circuitry, left unaddressed, result in a product performance that is insufficient to achieve its intended purpose. Some mechanism to reduce the impact of the various variations and imperfections is necessary to make that product viable.
Accordingly, various RF products often incorporate additional compensation elements to minimize the departure from the desired output profile. These compensation elements are often constructed by some additional RF components designed to cause an effect opposite to the undesired effects caused by the RF circuit and its components. Compensation of the gain variation over frequency is typically performed using RF equalizers consisting of resistors, capacitors and inductors. Compensation of gain variation over temperature is typically performed using a resistive attenuator, which incorporates an element with temperature varying resistance (i.e., thermistor) to change the attenuation as a function of temperature. Compensation of gain frequency-tilt variation over temperature is typically performed using an RF equalizer, while a thermistor is used to change the attenuation at a certain high frequencies while other components (such as capacitors, inductors, and/or resistors.) are used to lessen the attenuation at lower frequencies in a predetermined manner. Various other compensation circuit types exist to compensate for various other undesired RF effects introduced by imperfections in the RF components and its RF circuits that implement the main product desired functionality.
Unfortunately, all types of RF compensation circuits typically exhibit a certain insertion loss (negative gain). That insertion loss may be minimized in some conditions, such as extreme temperatures and at certain frequencies; however, the insertion loss will be even higher in other conditions (such as other frequencies, other temperatures, etc.). As the amount of compensation that is required increased, so does the loss exhibited by the compensation circuit. This loss typically is dealt with by adding more gain to the product (such as by an additional amplifier), which often results in higher power consumption and higher product cost, even on top of the compensation circuit cost itself.
The compensation circuits themselves need to be carefully designed and optimized. In some cases, a generic optimization of the compensation circuit is insufficient, and the compensation circuit must be individually tuned during the production of every unit. Such a process is quite time consuming, and thus increases the production cost substantially. Even so, compensation circuits typically enable limited success, and residual imperfections still impair the performance of the RF product.
In some prior art products, a radio frequency (RF) signal is digitally generated, then converted to analog by an analog-to-digital-converter (ADC), and then further processed by RF circuitry. In such cases, the imperfection introduced by the RF components and its circuit may be compensated for by digital compensation functions. FIG. 1 is a block diagram of such digital compensation functions according to the prior art. Such a solution is often preferred in performance, cost, and power efficiency when compared to RF compensation circuits. Moreover, automated calibration techniques can be introduced during the manufacturing of a product to individually optimize some parameters of the digital compensation functions for every manufactured unit.
A typical implementation of such a calibration scheme will employ a measurement device which measures and quantifies the RF signal imperfections at the product output. These measurements are processed by the calibration system to produce compensation parameters for the digital compensation functions implemented in the product. The compensation parameters are stored in a non-volatile memory incorporated in the product. Thereafter, the product uses the non-volatile stored parameters to digitally compensate for the RF circuit imperfections and produce less imperfect, more ideal output signal.